The k-tied Normal Distribution: A Compact Parameterization of Gaussian Mean Field Posteriors in Bayesian Neural Networks.
Jakub SwiatkowskiKevin RothBastiaan S. VeelingLinh TranJoshua V. DillonJasper SnoekStephan MandtTim SalimansRodolphe JenattonSebastian NowozinPublished in: ICML (2020)
Keyphrases
- normal distribution
- posterior distribution
- probability distribution
- neural network
- bayesian inference
- gaussian distribution
- random variables
- posterior probability
- latent variables
- mixture distribution
- probability density function
- markov networks
- standard deviation
- bayesian framework
- parameter estimation
- covariance matrix
- free energy
- maximum likelihood
- pattern recognition
- bayesian networks
- graphical models
- variational approximation
- markov chain monte carlo
- belief networks
- hyperparameters
- maximum a posteriori
- em algorithm
- expectation maximization
- probabilistic model
- prior knowledge
- prior information
- nearest neighbor
- data mining